What's the minimum viable ecosystem in IoT?

"Are you the mushroom or the mycelium?"

Unlike an MVP, which is largely focused on one company and one product, a minimum viable ecosystem (MVE) explicitly involves multiple companies - and potentially multiple products - requiring orchestration to reach 'minimum viable' level.

This piece is a conversation between Gaye Soykök, Head of Emerging Technologies at Legal & General, and Pilgrim Beart, CEO of IoT BI tool DevicePilot, with contributions from IoT pioneers, Usman Haque and Garrett Kinsman.

Can you introduce yourselves?

GS: I started with computer science, and it's not been so much about the technology for me, I was always looking for the "thing that made the change" - what it could be and how we can make it happen. I've worked in several different companies in the value chain such as R&D, product companies, agencies, and consultancy, and now my current company is a financial services provider, Legal & General.

PB: Similarly, I started in computer hardware and software at startups and I've always been interested in trying to predict what could come next - the market and technology drivers. Over the past 20 years, I've been starting companies as a way to bring ideas to the market faster.

What is an MVE?

PB: When I first read your brief but intriguing LinkedIn post (All you need is a Minimum Viable Ecosystem...), I understood that you were extending the concept of the Minimum Viable Product, and expanding to apply not just to a new product but to whole new ecosystems. I think it has great relevance to IoT and it resonated strongly with the experiences I shared in my Tedx talk. So how did you come across (or invent?) the idea of a Minimum Viable Ecosystem?

GS: I was thinking, "How do we create the value today in a more complex landscape? What should you do and what shouldn't you do, and therefore what do you need others to do?" As the complexity increases, design, implementation and structure become a bit vague, ironically. I think we might need to put our architecture hat on to create new things in this complex new landscape.

PB: By "architecture", do you mean technical architecture?

GS: No, I mean the value modules, like partners, capabilities, etc, and how these all fit together.

PB: What do you understand by "ecosystem"?

GS: I used to be an architect and architects follow a layered approach. I see:

Value (at the top)
People (organisations, society, governments)
Services, or my favourite concept - jobs-to-be-done (JTBD) e.g. emptying bins or caring for the elderly
Implementation at the bottom (technology, APIs, contracts, liability, etc)

What is viability then?

GS: Minimum Viable is application when we're trying to do something new. There's indeed a close analogy with MVP: we don't just create products anymore - a website, an app. Maybe it's my luck, but almost always these days it's more complicated than only a product and it takes severeal different players to deliver and create value. "Viable" means that it's enough to test an idea. And if the concept works not just technically but for adoption, then it's viable enough, so it takes off and grows by itself. It's an organic metaphor.

PB: Yes, that "taking off" is what many IoT use cases need right now. Often, the first application has to pay the costs, e.g. smart streetlights in a city have to pay for the network, but then subsequent applications in the same ecosystem can ride on the back of that investment and quickly get returns - so the ecosystem as a whole starts to take off. So, smart cities may need to some things to sponsor that first application in order to reap the ecosystem benefits more quickly.

What about bootstrapping the system?

PB: i was recently chatting to Nodle co-founder, Garret Kinsman, about how they're overcoming the ecosystem bootstrapping problem by using cryptocurrency to incentivize a global IoT network, to jump-start the network which, in turn, will attract IoT subscriptions. A cryptocurrency automates the billing process, but also allows Nodle to incentivize and automate network growth in "long tail" markets like Sri Lanka or Vietnam. Cryptocurrencies reduce the friction of international financial ecosystems, enabling network effects that couldn't be realized before.

Garret also gave the example of JIO, a network operator in India, who - when they launched their 4G network gave away free LTE connectivity to select users - destroyed competition and drove the whole country's internet usage up by 7x in a single quarter!

Kinsman told us, "Today, India's internet is literally 100x cheaper than the USA. We seek to leverage network effects to do the same for IoT." ($10 per GB on Google Fi vs $.0.10 per GB on Jio)

Kinsman thinks it's possible to even create network effects within network effects, thus amplifying a product's reach. A great example of this is the App Store, which leverages a network effect to enable others to create their own.

GS: That's a fantastic example. Another topical example where I think of the MVE concept is digital identities - shareable, portable and secure - so, for example, a financial digital identity.

Can you tell us a bit more about digital identities within MVE?

GS: Within my company, there's a complex internal ecosystem of finance businesses: pensions, investments, savings, retirement, life insurance, housing. If I have to test a solution around the finance ecosystem, it's the best place. We don't own a bank, so any bank is high on the list for me to complete my finance MVE. Also, if I want to test inter-company sharing, it's very likely that I should start my MVE inter-company.

If I want to test both finance and commercial digital identities, an external node such as a network operator or utility company should be in the MVE as well. Digital identity is something that would give value day-to-day for the user, so what you're testing is whether it's actually doing that for the user. Also, whether the sharing will work and create value enough to adopt and grow.

The whole USP of these concepts is that they're shared and work across multiple providers, so that's what you need to test: with an MVE, the network effect starts to happen.

PB: That reminds me of what I call "customer portfolio": with a new company, or new product or feature, you need to have sufficient use cases or example customers to prove the "genericness" of the thing and prove you haven't engineered far too narrow a use case, and that you're not stuck in a silo. Diversification reduces the risk of being misled by an outlier.

GS: Spot on - portfolio. The word "ecosystem" comes from nature: scaling without diversification is dangerous - plant a million identical trees, and when they fall sick it's all gone. Nature uses diversification guard against this.

That's a beautiful analogy.

GS: Blockchain is about distributed value as well. Nodes versus network. Until recently, we mostly looked at the nodes and their value - but it's the network that creates the value of the network. The analogy is actually mushrooms - are you the mushroom or the mycelium?

PB: Great analogy. Analogies with nature - the ultimate complex emergent system - can be helpful. I recently characterized IoT as Gaia's nervous system, and I wasn't just being cute - it seems that capabilities in systems tend to emerge when the system "needs" them. I recently read and enjoyed Peter Wohlleben's The Secret Life of Trees which is a fascinating study of how trees benefit from being an active and reciprocating part of their ecosystem.

When I discussed this with Usman Haque, it reminded him of his own work in designing participatory systems in general and how to bring about cooperation in a competitive environment. In trying to design for cooperation, even when everybody agrees on an end goal, it doesn't mean that any individual is going to take the first step - you actually need to design in strong, short term incentives for others, which to a certain extent means building trust.

When structuring participation, Usman's always tried to think about how the high cost of incentivising early participants can be justified by the low cost of those that they later bring in - and it's really important that incremental participation has incremental gains - i.e. if it's an "all or nothing" approach, then that's a strong disincentive to anyone who's even slowly sceptical. Having a "small reward" for "small participation" is a low risk proposition. Usman actually wrote some of this up in 2010.

So, in terms of process, do the concepts in MVP apply to MVE?

GS: I think so. It's about being efficient and accelerating learning. But could be even more fundamental than that - it's about avoiding cul-de-sacs by testing hypotheses early.

PB: When developing a product, early customers can be valuable - not so much for their revenue, but more for what they teach you. So pushing out a rough, early product allows you to make mistakes and learn faster.

GS: Yes, another way to think about that is that you're developing a robust immune system - you can survive setbacks, and tolerate (and even benefit from!) continued experimentation as you grow. Co-creation - another thing you mentioned in your Tedx talk, Pilgrim.... if someone is in a silo in a company, developing a proposition, they'll have challenges, but at least they're contained. But if they're building something across say, 12 companies, the complexity is much higher, especially if they're iterating quickly, so they need an immune system to help them tolerate failure.

The key idea is that it's got to work for everyone in the ecosystem, and it's very hard to do that by top-down planning, so MVE is definitely a bottom-up approach. If the ecosystem is healthy, then it survives and grows by itself, bottom-up.

Something I've become very interested in recently is "Rendanheyi " -

Wait, can you tell us what Rendanheyi is?

GS: - so the employee and the customers working together to disrupt the organisation for its sake. Create 1:1 value instantly and then evolve. Don't plan for scale, just do something useful, find some value to give, then repeat and grow. This doesn't necessarily scale linearly - you know, doing more of the same - because scale comes from doing diverse activities that support each other around the initial seed.

Haier - the test bed company of Rendanheyi - was just a white goods company, but with an evolutionary approach they became very successful. But it can be tough -t he organisation says "you can do whatever you want to... you just have to survive."

There's a bootstrapping problem - if it requires six entities to collaborate to deliver any value then how do you get started? Companies used to need to scale to become sticky, for example to sell dresses you had to manufacture 10,000 of them to get economies of scale. But once a supply chain becomes more dynamic, then you can create something on demand, in low volume, and then very quickly change it when demand changes.

The industrial revolution required investment in big, expensive machines to deliver economies of scale, which then had to deliver the scale to pay back the investment - lumbering, inherently slow. New economy is digital, therefore there's no "heavy iron" investment - and it's inherently agile.

One term that perhaps captures this idea is "economies of scope".

Let's get the exact definition of that first.

"plural noun: economies of scope

1. a proportionate saving gained by producing two or more distinct goods, when the cost of doing so is less than that of producing each separately."

GS: This impacts even old school physical products. So, for example, a luggage company recently changed themselves from producing different types of luggage to a lifestyle company, creating services such as concierge, transport, etc. around their physical products. So, diversity isn't just about robustness, it's actually about how you grow your business.

PB: Very interesting. Reminds me of evolutionary biologist Stuart Kauffman's idea of "the adjacent possible". The idea that disruption can actually be good for an ecosystem reminds me that I'm looking forward to reading Anna Lowenhaupt Tpsing's The Mushroom at the End of the World soon - it's about how a rare mushroom thrives amidst capitalist destruction. Once you have a good immune system, you can do change with disruption, which without the immune system could have killed you.

This has been a fascinating discussion, Gaye and Pilgrim, thank you so much for expounding your ideas on the minimum viable ecosystem. We look forward to continuing the discussion with our friends across the IoT ecosystem.

Interested in the references?

Ron Adner, 2012 - The Wide Lens. Potentially the first mention of the "Minimum Viable Ecosystem"
Dave Mather, 2012 - Innovating to create an ecosystem. A write up on Adner's talk.
Gaye Soykök, 2019 - All you need is a Minimum Viable Ecosystem. The original MVE post that sparked Pilgrim's curiosity.
Kevin Kelly, 1994 - Out Of Control: The New Biology of Machines, Social Systems, and the Economic World
Anna Lowenhaupt Tsing, 2015 - The Mushroom at the End of the World
Usman Haque, 2010 - Notes on the Design of Participatory Systems - for the City or for the Planet